Diagnostic T2 corrected

[xuelingbo]

The temperature of urban tile TH2_URB2D should be air temperature, not potential temperature. The diagnostic T2 calculation have been corrected.

LSP-DS/modified/module_NoahMP_hrldas_driver.F

Line 1078 in df6f93d

NoahmpIO%TH2_URB2D(I,J)*NoahmpIO%FRC_URB2D(I,J)

A disscussion about this issue is here: wrf-model/WRF#2093

[jenntibay]

Hi, Lingbo! Tested this new modification (corrected T2 computation):

V5.1
NoahmpIO%T2(I,J) = (NoahmpIO%FVEGXY(I,J)*NoahmpIO%T2MVXY(I,J) + (1.-NoahmpIO%FVEGXY(I,J))NoahmpIO%T2MBXY(I,J))(1.-NoahmpIO%FRC_URB2D(I,J)) + (NoahmpIO%TH2_URB2D(I,J)/((1.E5/NoahmpIO%FORCPLSM(I,J))**rcp))*NoahmpIO%FRC_URB2D(I,J)

V5.2
NoahmpIO%T2(I,J) = (NoahmpIO%FVEGXY(I,J)*NoahmpIO%T2MVXY(I,J) + (1.-NoahmpIO%FVEGXY(I,J))NoahmpIO%T2MBXY(I,J))(1.-NoahmpIO%FRC_URB2D(I,J)) + NoahmpIO%TH2_URB2D(I,J)*NoahmpIO%FRC_URB2D(I,J)

Attaching here figures of actual values at 2AM, 2PM, mean APRIL for T2, T2MV, TH2_URB2D, TEMP without urban, and the bias between two versions.

[xuelingbo]

Great, it looks good to me!

[xuelingbo]

I did a simulation using HRLDAS in Riyadh (capital city of Saudi Arabia), here are the plots of results:


Riyadh is a high altitude dessert city, no vegetation. As we can see from the figure, if we consider TH2_URB2D as potential temperature, there is no UHI, but if we consider TH2_URB2D as air temperature, the UHI shows.